Results suggest that difficulties in hot EF underlie externalizing problem behaviors in middle childhood.

Hassall 1983. 11 'Although the wall has been intensively studied for years in only four forts have barracks been completely excavated (Benwell, Halton Chesters, Housesteads and South Shields) ...' (Breeze and Dobson 1969, 25). The evidence was even more meagre than this statement allowed: at these four sites (with the possible exception of Housesteads) the barrack plans were not revealed in detail. 'Complete excavation' in the sense of an area excavation of all surviving deposits had yet to occur on the northern frontier.

Most students of Roman frontiers are familiar with the idea that, before the construction of Hadrian's Wall in c. 122, a system of frontier control had developed on the Tyne—Solway isthmus under Trajan, particularly after the withdrawal of the Roman army from southern Scotland in c. 105 (FIG. 1). A road running between Corbridge and Carlisle, known in medieval times as the Stanegate, has generally been seen as the basic component of this Trajanic frontier. The idea of the Stanegate road as a frontier originated with Forster, was developed by Collingwood, and the schedule of sites along it was formalised by Birley, who proposed a regular Stanegate system of forts at intervals of half a day's march, alternating with fortlets.

I t is generally held that Hadrian's Wall was abandoned c. A.D. 140 in favour of a heavily garrisoned Antonine Wall, then reoccupied about A.D. 158 upon withdrawal of troops from Scotland and the Antonine Wall. Almost immediately troops returned to Scotland, though the northern garrison was thinner than before. This second, brief occupation of the Antonine Wall is now most often supposed to have ended at some date early in the reign of Marcus Aurelius, Hadrian's Wall then being reoccupied. The basic evidence for the theory of two distinct Antonine occupations of the Antonine Wall, and Scotland in general, is epigraphic and archaeological. It is commonly stated that, where excavated, most Antonine sites in Scotland have displayed two periods. At several sites inscriptions record two successive Antonine garrisons. The purpose of this paper is to reexamine the structural and epigraphic evidence in order to establish whether belief in two distinct Antonine periods in Scotland is really warranted. 1 If the Antonine occupation of Scotland could be shown to be of a single period, new possibilities would be raised for the date of the final abandonment of the Antonine Wall, a question to which the definitive answer seemed to have been given in Britannia over twenty years ago. 2 A different light would also be cast upon certain assumptions, now quite deeply embedded in the literature on the subject, about the ease with which the Romans entered Scotland and the nature of the events that caused them finally to relinquish it. The appendix (p. 43) presents the evidence for two Antonine periods at every excavated Antonine site in Scotland, and should be consulted for bibliographical details of sites discussed in the main text. It will be clear from the appendix that the belief that the majority of excavated forts in Scotland show two distinct periods of Antonine occupation is false. Until the 1960s, Macdonald's view, proposed in the second edition of his Roman Wall in Scotland (1934), that three distinct periods were clearly visible on Antonine Wall sites, prevailed. These were dated 142-55, 158-180 and c. 184-5. K.A. Steer showed that 1 I record my thanks to Dr D.A. Welsby, who independently arrived at a critical view of Antonine I and II, and with whom it was originally intended to write a joint paper. While grateful to Dr Welsby for discussion, I am solely responsible for the present paper, the views expressed, and for any errors. I would also like to thank the following, who have read or commented on various versions of this paper, and offered support and constructive discussion:

No abstract